Microbial nitrogen limitation in the mammalian large intestine

Transcription

1 SUPPLEMENTARY INFORMATION Articles In the format provided by the authors and unedited. Microbial nitrogen limitation in the mammalian large intestine Aspen T. Reese 1,2, Fátima C. Pereira 3, Arno Schintlmeister 3,, David Berry 3, Michael Wagner 3,, Laura P. Hale, Anchi Wu 2, Sharon Jiang 2, Heather K. Durand 2, Xiyou Zhou 6, Richard T. Premont 6, Anna Mae Diehl 2,6, Thomas M. O Connell 7, Susan C. Alberts 1,8,9, Tyler R. Kartzinel, Robert M. Pringle 11, Robert R. Dunn 12, Justin P. Wright 1 and Lawrence A. David 2,13 * 1 Department of Biology, Duke University, Durham, NC, USA. 2 Department of Molecular Genetics and Microbiology, Duke University, Durham, NC, USA. 3 Department of Microbiology and Ecosystem Science, Division of Microbial Ecology, Research Network Chemistry Meets Microbiology, University of Vienna, Vienna, Austria. Large-Instrument Facility for Advanced Isotope Research, Research Network Chemistry Meets Microbiology, University of Vienna, Vienna, Austria. Department of Pathology, Duke University Medical Center, Durham, NC, USA. 6 Department of Medicine, Duke University Medical Center, Durham, NC, USA. 7 Department of Otolaryngology Head & Neck Surgery, Indiana University School of Medicine, Indianapolis, IN, USA. 8 Department of Evolutionary Anthropology, Duke University, Durham, NC, USA. 9 Institute of Primate Research, National Museums of Kenya, Nairobi, Kenya. Department of Ecology and Evolutionary Biology, Brown University, Providence, RI, USA. 11 Department of Ecology and Evolutionary Biology, Princeton University, Princeton, NJ, USA. 12 Department of Applied Ecology, North Carolina State University, Raleigh, NC, USA. 13 Center for Genomic and Computational Biology, Duke University, Durham, NC, USA. * lawrence.david@duke.edu Nature Microbiology

2 a Fecal Day b Microbial load copies Copies 16S rrna 16S/gram gene/gram feces feces 1e+9 1e+8 1e+7 3 Diet Low Control High c Bacteroidaceae Copies 16S/ Bacteroidaceae relative abundance gram feces High Control Low Protein Content d Bacteroidaceae Copies load 16S/ copies 16S gram rrna gene/gram feces feces 1e+8 1e+6 3 f Weight (g) Day Supplementary Figure 1 Changing dietary protein content results in changes in associated with some microbial shifts. a, Fecal changed under altered protein diets over 1 days (P< 2.2x -16, linear mixed effects model likelihood tests; n=9- mice per treatment group). b, Microbial load, estimated by 16S rrna gene copy number measured via qpcr, was negatively correlated (linear regression shown) with fecal on the final day of treatment (ρ=-., P=.3, Spearman correlation; n=28 mice) c, Bacteroidaceae relative abundance did not differ between treatment groups (P=.8, Kruskal-Wallis test; n=9- mice per treatment group). d, However, Bacteroidaceae absolute abundance (calculated as relateve abundance multiplied by total copies 16S rrna gene measured via qpcr) trended towards a negatively correlation (linear regression shown) with fecal on the last day of treatment (ρ=-.29, P=.1, Spearman correlation; n=28 mice). f, Mouse weight did not differ significantly between diet treatments (P=.93, Kruskal-Wallis test; n=9- mice per group). Large circles are means; bars show standard deviations.

3 a b 1 16 * *** Section Proximal Small Intestine Distal Small Intestine Large Intestine 8 6 Proximal Proximal Distal Distal Small IntestineSmall Intestine Small Intestine Small Intestine Large Large Intestine Intestine 1e+6 1e+7 1e+8 16s Copies/Gram Microbial dried load feces copies 16S rrna gene/gram dried gut contents Supplementary Figure 2 of digesta varies along the length of the gut but is not correlated with total microbial load. a, increased along the gut of mice under normal diets (P=. proximal small intestine compared to distal small intestine, P=1 proximal small intestine compared to large intestine, Tukey s Honestly Significant Difference test; n= mice per compartment). Bars indicate groups which differed significantly (*** P<1, * P<.). b, Microbial load, estimated by 16S rrna gene copy number, was not significantly correlated with gut content (ρ=.2, P=.1, Spearman correlation; n= per compartment) when analyzed across the entire mouse intestine. Large circles are means; bars show standard deviations.

4 c Thickness (um) b 1e+8 1e+7 1e+6 Muc2/Actb expression Microbial load copies 16S16s/Gram gene/gram feces Copies Feces a 1 3 d 1.7 Control Treated abx Antibiotics Control Antibiotics Day 1.2 Supplementary Figure 3 Antibiotic treatment affects gut microbial load and mucus thickness. a, Antibiotics produce a significant decrease in microbial loadduring the day treatment course 1 1 (P=9.1x-6, linear mixed effects model likelihood test; n=9- mice per treatment group)), after which microbial load converged between treated and un-treated mice. Microbial load was measured as 16S rrna gene copy number per gram feces extracted with qpcr detection limits between ~^ and ~^9. b-d, thickness was measured on Carnoy s fixed mouse gut sections stained with Alcian blue after days of treatment. Blinded measurements of an intact AB-stained layer immediately overlying the colonic epithelium, which we believe based on previous literature to represent the inner mucus layer, revealed a trend toward thinner mucus in antibiotics treated mice (P=.9, Mann-Whitney U test; n=9 mice per treatment group; b). However, we did not attempt to measure the thickness of the more amorphous and bacterial-laden outer mucus layer. Representative images for control (c) and antibiotic treated (d) mice include a μm scale bar. Images were prepared for all samples (N=18). See Supplementary Information Table 7 for individual measurements of mucus thickness. Red bar under the x-axis indicate the course of antibiotics. Large circles are means; bars show standard deviations.

5 estine s estine s um elium m um es es s s estine s estine s a ntestine cus ntestine cus Relative Nitrogen nitrogen Allocation allocation (At% excess (At% Excess 1 N/total N/Total 1 1 N administered (g)) N Administered) APE_n_in APE_n_in APE_n_in hr 6hr hr 6hr hr 6hr Small Intestine b c Small Intestine Large Intestine Feces Small Small Intestine Large Intestine Feces Epithelium Feces Feces Epithelium Epithelium Epithelium Epithelium hr 6hr hr 6hr hr 6hr as.factor(hour) hr 6hr hr 6hr hr 6hr Epithelium Epithelium Epithelium Epithelium Feces d Epithelium 6 6 Epithelium 6 6 Epithelium 6 6 Feces Feces Feces Feces 6 6 Epithelium Feces Feces Feces Epithelium Epithelium d Small Intestine Large Intestine 16 1 Small Intestine Large Intestine Control as.factor(hour) as.factor(hour) 1 1Secreted Dietary hr 6hr hr Feces 6hr hr 6hr Feces hr Epithelium Feces 6hr hr Epithelium Feces 6hr hr Epithelium 6hr Large Intestine Large Feces Feces Epithelium 12 Feces Epithelium Epithelium Feces Epithelium Epithelium Feces Epithelium Feces Epithelium Epithelium Epithelium Epithelium Epithelium Epithelium 6 Epithelium Epithelium 6 Epithelium Epithelium 6 FecesEpithelium Epithelium Feces Mucu Feces Feces Epithelium Epithelium Epithelium Epithelium Epithelium Epithelium Mucu Large Intestine Small Intestine Epithelium Epithelium Epithelium Epithelium Epithelium Epithelium Epithelium Epithelium Epithelium hr 6hr hr 6hr hr 6hr hr 6hr Epithelium hr 6hr hr Epithelium 6hr hr 6hr hr 6hr 1 1 Epithelium Supplementary Figure Labeled nitrogen from both host diet and host secretions was delivered to the gut and made available to the microbiota without changing nitrogen availability. a, Isotopic enrichment delivery Epithelium to the gut was demonstrated throughout the gut after four and six hours. Bars indicate difference between treatments (P<., Bonferroni-corrected Mann-Whitney U tests (see Supplementary Information Table 8 for exact p-values); n=6 per treatment group). b, did not vary between treatments (P=.13, Bonferroni-corrected Kruskal-Wallis test), indicating the treatments did not change overall nitrogen availability. c-d, However, does vary between compartment (small intestine, cecum, large intestine, feces) and among tissue layers within each compartment (epithelium, mucus, lumen) (P<., Bonferronicorrected Kruskal-Wallis tests (see Supplementary Information Table 9 for exact p-values); n=12-18 per section) after (c) and 6 (d) hours. Large circles are means; bars show standard deviations. Boxplots indicate medians and quartiles; whiskers show 1.*interquartile range. APE_n_in APE_n_in as.factor(hour) APE_n_in Epithelium Epithelium Epithelium Epithelium Epithelium Epithelium

6 .7 Small Intestine Large Intestine Small Intestine Large Intestine a b c Fecal 12 At% 1 N. At% 1 N. Group At% 1 N Conventional.38 Germfree.37 Group Conventional Germfree Group Conventional Germfree 8.. Conventional Germfree Epithelium Epithelium Epithelium Epithelium Epithelium Secreted Threonine Epithelium Epithelium Epithelium Supplementary Figure Nitrogen allocation patterns in germfree mice. a, Fecal did not differ between conventional and germfree mice (P=.9, Mann-Whitney U test; n=8- mice per group). b-c, Atom percent enrichment (i.e., the proportional representation of the heavy isotope times ) four hours following stable isotope tracer delivery via injection (i.e., host secreted; b) or host diet (c) differed between germfree and conventional mice (P=7 and P=.261, respectively, mixed-model ANOVAs; n=6 per treatment group). Large circles are means; bars indicate standard deviation. Boxplots indicate medians and quartiles; whiskers show 1.*interquartile range. Threonine Epithelium Epithelium.36 Epithelium Epithelium Epithelium Dietary Epithelium Threonine Ep

7 a Secreted b Dietary c FISH DAPI d at% 1 N e f at% 1 N at% 1 N 3 Supplementary Figure 6 Representative NanoSIMS/FISH analysis of large intestine microbiota from mice fed a 1 N labeled spirulina diet or injected with 1 N/ 13 C labeled threonine. a-d, Large intestine contents were stained with the nucleic acid dye DAPI (gray) and hybridized with the FISH Bac33 (green) and Erec82 (red) probes to visualize bacterial cells of particular taxa (see Supplementary Table 3). e-f, NanoSIMS at% 1 N distribution maps of the same fields of view for secreted 1 N (e) and dietary 1 N (f) are shown. Yellow arrowheads point to higher enrichment levels detected for Bac33 positive cells in comparison with Erec82 positive cells, which are indicated by white arrowheads. All measured bacterial cells are outlined in grey, for easier visualization. Grey background areas (e-f) refer to pixels where an unbiased at% visualization is not feasible due to weak CN- secondary ion signal intensities (ITO surface without biomass). Scale bar = μm. Images were prepared for all samples prepared for NanoSIMS (n=6).

8 a injection fast spirulina Hours b Microbial labeling At% 1 N Cell (At% 1 N cell) 1 At% 1 N Unenriched Unenriched Enriched Enriched 1 1 NN 1 N Enriched Enriched CC 1 1 N 13 NC 1 N 1 c Bulk labeling (At% At% 1 1 N N) Control Secreted Dietary 1 Compartment LI Large Intestine SI Small Intestine Clostridium lusters XIVa/XIVb d XIVa/XIVb NMDS Compartment LI Large Intestine SI Small Intestine Control Secreted Dietary 16 Treatment Control Spirulina Threonine 12 e 16 Large Intestine 12 Control Secreted Dietary Small Intestine Control Secreted Dietary NMDS1 Large Intestine Small Intestine Supplementary Figure 7 Labeled nitrogen from both host diet and host secretions did not change nitrogen availability or composition but was taken up by the microbiota. a, Experimental setup for single cell studies: mice were fasted overnight and then offered chow with all protein sourced from Spirulina cells (either 1 N for control and secreted groups or 1 N for dietary group) for one hour before being returned to normal mouse chow for four hours before euthanasia. Control and secreted-group mice received an injection of threonine ( 1 N and 1 N respectively) before being returned to normal chow. b-c, The diet manipulation produced the greatest overall 1 N enrichment of microbial cells (P<2.2x -16, Bonferroni-corrected Mann-Whitney U tests; n= 2 mice per treatment, n= cells per treatment, b) likely due to much greater label uptake during feeding, but both interventions resulted in significant overall enrichment relative to controls with no significant differences between intestinal compartments (colors; P=.16, Mann-Whitney U test, n=- mice per treatment; c). d, The experimental treatments (colors) did not significantly affect microbial taxonomic composition (P=.9) as assessed by 16S rrna gene amplicon sequencing but taxonomic composition did differ significantly between gut compartments (shapes; P=1, PERMANOVA; n=- mice per treatment). e, The experimental treatments did not affect gut content (P=.9), but did vary between intestinal compartments (P=.2, Mann-Whitney U test; n=- mice per treatment). Large circles are means; bars show standard deviations.

9 Supplementary Table 1: Gut isolate. For isolates with more than one clonal population measured (N>1) reported is the mean. Species Phylum N Source Collinsella aerofaciens Actinobacteria 1 Human donor.1 Eggerthella centa Actinobacteria 1 Human donor 3.99 Bacteroides massiliensis Bacteroidetes 1 Human donor.8 Bacteroides ovatus Bacteroidetes Human donor 3.9 Bacteroides thetaiotaomicron Bacteroidetes 1 Human donor.28 Bacteroides uniformis Bacteroidetes 1 Human donor.7 Bacteroides vulgatus Bacteroidetes 1 Human donor.22 Bacillus coccae Firmicutes 1 Human donor.33 Clostridium bartlettii Firmicutes 1 Human donor 3.8 Clostridium innocuum Firmicutes 1 Human donor.2 Clostridium innocuum Firmicutes 1 Human donor 3.8 Clostridium sp. Firmicutes 1 Human donor.17 Clostridium sp. Firmicutes 1 Human donor. Clostridium sp. Firmicutes 1 Human donor 3.6 Clostridium sp. Firmicutes 1 Human donor.31 Coprococcus comes Firmicutes 1 Human donor 3.98 Dorea formicigenens Firmicutes 1 Human donor.1 Dorea formigamerans Firmicutes 1 Human donor.1 Dorea longicatena Firmicutes 1 ATCC 3.97 Dorea longicatena Firmicutes 1 Human donor 3.78 Enterococcus faecalis Firmicutes Human donor 3.9 Enterococcus faecalis Firmicutes 1 Human donor.28 Enterococcus faecalis Firmicutes 1 Human donor 3.88 Ruminococcus gnavus Firmicutes 1 Human donor 3.76 Ruminococcus gnavus Firmicutes 1 Human donor 3.92 Ruminococcus productus Firmicutes 1 Human donor.2 Ruminococcus torques Firmicutes 1 Human donor 3.62 Streptococcus salivarius Firmicutes 1 Human donor 3.98 Streptococcus salivarius Firmicutes 1 Human donor 3.8 Streptococcus salivarius Firmicutes 1 Human donor 3.97 Enterobacter cloacae Proteobacteria 1 Human donor.2 Raoultella ornithinolytica Proteobacteria 1 Human donor.28 Raoultella ornithinolytica Proteobacteria 1 Human donor.12 Shigella flexneri Proteobacteria 1 Human donor 3.8 Akkermansia municiphila Verrucomicrobia ATCC.9

10 Supplementary Table 2: Mammals included in fecal analyses Large intestine Length Collection Common name Species N Diet Gut morphology length (cm) source locale Population Dog Canis lupus familiaris Carnivore simple 68 USA Captive Aye-aye Daubentonia madagascariensis Carnivore simple USA Captive White-tailed Mongoose Ichneumia albicauda 1 Carnivore simple Kenya Wild Aardwolf Proteles cristatus 1 Carnivore simple Kenya Wild Meerkat Suricata suricatta 9 Carnivore simple South Africa Wild Impala Aepyceros melampus Herbivore hindgut Kenya Wild White Rhinoceros Ceratotherium simum Herbivore hindgut Kenya Wild Black Rhinoceros Diceros bicornis 7 Herbivore hindgut.73 2 Kenya Wild Horse Equus ferus caballus Herbivore hindgut 77 USA Captive Grevy's Zebra Equus grevyi 9 Herbivore hindgut Kenya Wild Plains Zebra Equus quagga Herbivore hindgut Kenya Wild Crested Porcupine Hystrix cristata 2 Herbivore hindgut Kenya Wild Snowshoe hare Lepus americanus 1 Herbivore hindgut USA Captive Elephant Loxodonta africana Herbivore hindgut Kenya Wild Prairie Vole Microtus ochrogaster Herbivore hindgut 2. 2 USA Captive Rock Hyrax Procavia habessinica 2 Herbivore hindgut 7 2 Kenya Wild Hippopotamus Hippopotamus amphibius Herbivore pseudo-ruminant Kenya Wild Cattle Bos indicus Herbivore ruminant 16 Kenya Captive Cow Bos taurus Herbivore ruminant USA Captive Giraffe Giraffa cameloparadalis Herbivore ruminant Kenya Wild Waterbuck Kobus ellipsiprymnus 1 Herbivore ruminant Kenya Wild Gunther's Dik-dik Madoqua guentheri Herbivore ruminant Kenya Wild Sheep Ovis aries Herbivore ruminant USA Captive Cape Buffalo Syncerus caffer Herbivore ruminant Kenya Wild Mouse Mus musculus Omnivore hindgut 7.1 USA Captive Warthog Phacochoerus africanus 6 Omnivore hindgut Kenya Wild Vervet Monkey Cercopithecus pygerythrus 1 Omnivore simple Kenya Wild Human Homo sapiens Omnivore simple USA Wild Grey Mouse Lemur Microcebus murinus 8 Omnivore simple USA Captive Yellow Baboon Papio cynocephalus 8 Omnivore simple Kenya Wild

11 Supplementary Table 3: FISH probes for single cell stable isotope analyses Major target taxa (coverage %, total hits in Probe Name Probe sequence ( 3 ) Total hits a Total non-target hits taxon) a a References Order Clostridiales (32.2%, 9787) Erec82 GCT TCT TAG TCA RGT ACC G Family Lachnospiraceae (7.2%, 932) Order Bacteroidales (.9%, 861) Family Bacteroidaceae (91.%, 68) Bac33 CCA ATG TGG GGG ACC TT Family Porphyromonadaceae (17.1%, 82) Family Prevotellaceae (71.1%, 12) a According to RDP probe match, performed with database release 11, Update (26 December 1), containing 3226 bacterial and archeal 16S rrna sequences ( Coverage is the percentage of sequences within the RDP target taxon that shows a full match to the probe sequence. The number of nontarget hits indicates the total number of sequences outside the respective RDP taxa that show a full match to the probe sequence. 71

12 Supplementary Table : Genus level absolute abundance (calculated as relative abundance values multiplied by 16S copy number measured via qpcr) responses to dietary intervention at the end of two weeks. Uncorrected and Bonferroni-corrected p-values are reported for Kruskal-Wallis tests of treatment effects. Only genera with an average abundance of at least.% and at least one sample with greater than 1% abundance were included. 6% Protein Diet % Protein Diet % Protein Diet adjusted Phylum Order Genus mean SD mean SD mean SD p-value p-value Bacteroidetes Bacteroidales Bacteroides.8E+6 1.E E+7 2.7E+7.E+7.E+7. 1 Parabacteroides 1.9E E+7.18E+7.6E+7.E+7.E Rickenellaceae genus unspecified 1.8E+ 2.99E+ 6.2E+.28E+ 6.E+.7E S2-7 genus unspecified 1.11E E+7 1.3E E+7 9.6E E Deferribacteres Deferribacterales Mucispirillum 7.23E+ 1.26E+6 1.2E+6 1.8E+6.87E+ 8.29E Firmicutes Lactobacillales Lactobacillus 3.21E+ 6.8E+.93E+6 1.E E E+7 9 Lactococcus 3.E+ 2.82E+ 1.1E+6 1.9E+6 6.3E E Clostridiales Family unspecified 7.79E+6 9.E+6 1.E+7 1.2E E+7 1.3E SMB3 1.2E+6 9.E+ 2.6E+6.11E E+6.37E Lachnospiraceae genus unspecified 1.3E E+6 1.7E E+6 2.E+6 2.7E Dorea.17E+.28E+.8E+ 6.2E+.63E+6 6.1E rc- 7.37E E+7 9.7E+6 8.8E+6 2.9E E Ruminococcaceae genus unspecified 2.77E E+6.61E+6.E E+6.7E Oscillospira 2.1E E+6 1.9E E E+7 9.E Ruminococcus 6.E+ 6.33E+ 2.26E+6 2.E E E Erysipelotrichales Erysipelotrichaceae genus unspecified 3.2E+.3E+.1E+ 2.99E+ 1.E E [Eubacterium] 2.E+ 3.79E+.3E+ 1.3E E+6 9.2E Tenericutes RF39 Family unspecified.21e+ 9.81E+ 1.8E+6 1.2E E+6 2.3E Verrucomicrobia Verrucomicrobiales Akkermansia.22E+6.8E+6 1.1E E+7.13E+7 7.E

13 Supplementary Table 6: Select nutrient data for experimental mouse diets including primary protein and carbohydrate sources. Diet Name Protein (% by weight) Casein (g/kg) Carbohydrate (% by weight) Sucrose (g/kg) Cellulose (g/kg) Fat (% by weight) Calcium (% by weight) Phosphorous (% by weight) Kcal/g TD TD TD

14 Supplementary Table 7: Mean mucus thickness measurements for antibiotics experiment data presented in Extended Data 3b. Mouse Antibiotics Thickness (um) ± SD Measurements (N) 1 N.8± N 7.8± N 8.9± N 6.9± N 16.± 6. 6 N 21.± N 9.1± N 8.2± N 23.8± Y 37.7± Y.9±. 3 Y 6.7± Y 9.1± 2. 6 Y.± Y 6.6± Y 7.8±. 8 Y.9± Y 6.1± 2.8 9

15 Supplementary Table 8: Bonferroni-corrected p- values for Mann-Whitney U tests to determine effect of labeling route (secreted or dietary) on isotopic enrichment in the gut. Significant effects in bold. Plots of enrichment values are in Supplementary Information Figure a. Hour Layer Compartment Treatment effect p-value Epithelium Small Intestine.217 Small Intestine.217 Small Intestine.33 6 Epithelium Small Intestine Small Intestine Small Intestine.217 Epithelium Epithelium Epithelium Large Intestine.217 Large Intestine 1 Large Intestine 1 6 Epithelium Large Intestine Large Intestine. 6 Large Intestine 1 Feces 1 6 Feces 1

16 Supplementary Table 9: Bonferronicorrected p-values for Kruskal-Wallis tests to determine effect of gut compartment and layer on during i s o t o p i c e n r i c h m e n t s t u d i e s. Significant effects in bold. Plots of are in Supplementary Information Figure c, d. Hour Effect p-value Compartment 6 Layer 8 Layer (small intestine).718 Layer (cecum) Layer (large intestine) 7 6 Compartment 83 6 Layer 29 6 Layer (small intestine) 6 Layer (cecum) Layer (large intestine) 27